The recent anthropogenic global warming makes a detailed knowledge of variations in the Earth climate system and of the coupling processes between climate and biogeochemical cycles of pressing importance. Studies of climate changes in the past represent a vital part of climate change research which is essential to assess the current warming against the background of natural climate variability. Due to strong limitations in direct observations, climate reconstructions for the past can only be achieved using natural climate archives. The paleoclimatic archive in ice cores provides not only information on climate variability over many thousands of years in high resolution but also on greenhouse gases, aerosol concentrations and more. Crucial questions on climate variability on interannual to orbital time scales and on the coupling processes and teleconnections in the climate system remain still open. To answer these questions novel climate parameters on polar ice cores are needed that go beyond previous studies in terms of temporal resolution, spatial coverage as well as quantitative representativeness. This proposal intends to develop such methods based on latest advances in analytical techniques and to apply them to polar ice cores. The common theme of the new approaches within MATRICs is the reconstruction of new, quantitative temperature information from different regions of the Earth all on the same core avoiding crucial crossdating issues. This comprises (i) continuous quantitative reconstructions of local temperature changes on polar ice sheets in seasonal resolution using new approaches, (ii) estimates of climate changes in continental, not permanently ice covered regions based on concurrent changes in the methane cycle and (iii) a new physical ice core gas thermometer for mean global ocean temperature. Successful implementation of the studies in MATRICs will make a significant contribution to maintain the world leading position of European ice core science.